KR20050057040A - Sound synthesiser - Google Patents

Abstract

A sound synthesiser is provided that reduces the computational requirements of a synthesiser with a high degree of polyphony, while ensuring that audible artefacts are kept to a minimum. The synthesiser comprises a plurality of samples stored in a memory (6) ; a plurality of voices (4, 16) each comprising means for calculating an output using a plurality of samples selected from the plurality of samples stored in the memory (6); wherein a voice (4, 16) is active when calculating an output; wherein the number of samples selected by the means for calculating depends upon the number of active voices (4, 16).

Description

Sound synthesizer {SOUND SYNTHESISER}

The present invention relates to sound synthesizers, in particular to sound synthesizers used in devices with limited computational resources, such as in portable devices.

Modern sound synthesizers are required to have many voices. The number of voices a synthesizer has is limited by the number of sounds that can be produced simultaneously.

There are several other protocols and standards that specify how an electronic sound synthesizer reproduces a set of required sounds.

One popular way to generate sound in electronic devices is by using the MIDI (Electronic Instrument Digital Interface) protocol. Unlike digital audio files, MIDI files (such as those found on compact discs) do not contain detailed descriptions of specific sounds. Instead, the MIDI file contains a list of events that the device must perform to recreate the correct sound. The sampled sound is stored in the synthesizer and used according to the instructions contained in the MIDI file. Therefore, MIDI files can be much smaller than digital audio files, making them suitable for environments with limited storage memory.

In the General MIDI System Level 1 (GM-1), the synthesizer is required to have at least 24 voices.

Synthesizers, such as MIDI synthesizers that generate sound from pre-recorded sound, are known as wave-table based synthesizers. In such a synthesizer, one or several pre-recorded sequences of the electronic musical instrument will be stored in a wave-table. Each sequence will contain consecutive samples played to recreate the sound.

Often, electronic musical instruments can generate very many notes, so sampling and recording all possible notes requires a lot of memory, so only a few notes are stored.

Therefore, if a synthesizer is required to produce a note or sound with a different frequency than that stored in memory, the synthesizer uses one of the stored sequences, re-samples it and changes the frequency to obtain the requested tone. Use a technique known as 'sample rate conversion'.

Changing the frequency of a stored sequence is accomplished by using the stored samples at various rates. That is, for example, if stored samples represent musical notes at a frequency of 300 Hz, then using all samples in sequence means reproducing the musical notes at 300 Hz. If each stored sample is output twice before the next stored sample is read, the note reproduced by the synthesizer will have a frequency of 150 Hz. Similarly, if a 600 Hz tone is required, the stored sample is read every other time.

It is important that the rate at which the sample is output by the synthesizer is constant and equal to one sample period (the time between each stored sample).

In this example, by using every sample twice, artifacts (distortions) will occur in the output sound. To overcome this distortion, the synthesizer calculates additional samples based on the stored samples. Therefore, in the 150 Hz example above, instead of repeating each stored sample twice, the synthesizer will output one stored sample and calculate the next sample based on the surrounding stored samples.

To do this, the synthesizer requires an interpolation technique.

The simplest interpolation technique uses a weighted average of two surrounding samples. However, this technique is often inaccurate and still cause audible distortion.

The optimal interpolation algorithm uses the sin (x) / x function and requires an infinite number of calculations. Of course, this is not practical, so sub-optimal algorithms have been developed.

One sub-optimal interpolation technique is described in Chapter 8 of "Applications of DSP to Audio and Acoustics" by Dana C. Massie. In the interpolation technique, several stored samples are calculated in the calculation (the number of samples used in interpolation is Known as the degree). The greater the number of samples used in interpolation, the better the synthesizer's performance.

In synthesizers, each voice is implemented using one or several digital signal processing (DSP) and the computational power of the DSP system limits the number of voices that the synthesizer can make and also uses for each voice. Limit the interpolation degree that is enabled

When using sub-optimal interpolation algorithms, such as the truncated sin (x) / x algorithm, the computational complexity increases linearly due to the interpolation degree.

In many commercial synthesizers, ten interpolation degrees are often used because they cause a good trade-off between computational complexity and sound quality.

1 is a diagram illustrating a sound synthesizer according to the present invention.

2 illustrates a method performed by the controller of FIG. 1 in accordance with the present invention.

3 is a diagram illustrating a method of determining an interpolation degree based on the number of active voices according to the present invention.

4 illustrates an alternative way of determining interpolation degree based on the number of active voices in accordance with the present invention.

5 illustrates in more detail the voice of the synthesizer for FIG. 1;

6 illustrates a mobile phone including a music synthesizer in accordance with the present invention.

It is desirable to be able to implement a MIDI sound synthesizer in a portable device such as a mobile phone, and to be able to produce polyphonic ringtones and high quality sounds in the device.

However, the limitations on computing power of portable devices (such as the cost and available space on the device) are in accordance with general MIDI system level 1 (GM-1) (i.e. with 24 voices) and about 10 interpolation degrees. Is not sufficient to implement a sound synthesizer.

Therefore, the present invention seeks to provide a sound synthesizer that reduces the computational requirements of synthesizers with high degree of polyphony while keeping the audible artifacts to a minimum.

Therefore, according to a first aspect of the invention, there is provided a memory comprising a plurality of stored samples; A synthesizer is provided that includes means for calculating an output signal for each of the plurality of active voices, using the plurality of samples selected from the stored samples for each of the plurality of active voices; By means of the calculating means, the number of samples used for each active voice depends on the number of active voices.

It is desirable for each voice to calculate only one output at a time.

By means of the calculation, the number of samples used for each active voice is preferably reduced as the number of active voices increases.

By means of the calculation, the number of samples used for each active voice is preferably reduced as the number of active voices increases so as not to exceed the maximum computational complexity.

Alternatively, by means of the calculation, the number of samples used for each active voice decreases non-linearly as the number of active voices increases.

The plurality of samples stored in the memory preferably include samples of sheet music.

The plurality of samples stored in the memory preferably include samples of sheet music made by various electronic musical instruments.

According to a second aspect of the invention, there is provided a portable device comprising a music synthesizer as described above.

The portable device is preferably a mobile phone.

Alternatively, the portable device is a wireless pager.

When the term "comprising / includes" is used herein to specify a feature, integration, step, or component presently mentioned, it is emphasized that this does not exclude other features, integration, steps, components or combinations thereof. Lose.

Reference is made to the examples for the accompanying drawings in order to illustrate a good understanding of the invention and how to practice it.

1 shows a music synthesizer in accordance with the present invention. As conventionally, the synthesizer includes a controller 2, a plurality of voices 4, a wave-table memory 6, a filter table 8, a mixer 10 and a digital-to-analog conversion module 12. do.

The synthesizer is described below as a wave-table based synthesizer using the MIDI protocol, but the present invention will be appreciated as suitable for any wave-table based synthesizer required to calculate a sample between two stored samples.

It should be noted that the term 'sample' as used below relates to a single audio sample point.

The total number N of voices 4 in the synthesizer defines the maximum polyphony of the system. By increasing N, polyphony increases by increasing the number of sounds to be made at the same time. For MIDI synthesizers that conform to General MIDI System Level 1 (GM-1), the value of N will be at least 24. Briefly, only three voices are shown in FIG. 1.

The controller 2 receives data via input 14. The data will include a stream of MIDI information for a portion of the music or for a particular series of sounds. Each MIDI file will contain a list of events describing the specific steps the synthesizer must perform to produce the required sound.

In the case of a MIDI file stored in a portable communication device, the file would be for a short piece of music that could be used, for example, as a link tone.

The controller 2 processes the MIDI data stream and sends the appropriate portion of the data to the relevant voice 4 so that the required sound can be synthesized. For example, since the required sound consists of several different instruments played at the same time, each voice 4 will simultaneously process a portion of one monophonic or polyphonic instrument. Often, the MIDI file will contain commands associated with the particular voice 4 to be used in synthesizing the next output.

Depending on the specific content of the MIDI file, different numbers of voices will be used at some time, depending on the particular part of the music to be played.

Each voice 4 is connected to a controller 2, a mixer 10, a wave-table memory 6 and a filter table 8.

Wave-table memory 6 includes a sequence of multiple digital samples. Each sequence will, for example, represent a score for a particular electronic musical instrument. Due to memory limitations, only a few notes will be stored per instrument.

The filter table 8 contains a number of filter values. In a preferred embodiment, the value represents a sinc function (where the sinc function is sinc (x) / x).

Although not shown in FIG. 1, both the wave table memory 6 and the filter table 8 are each limited to a sample period (the sample period is limited to the inverse of the sampling rate, that is, the rate at which the original sound is sampled). It contains a multiplexer that allows it to be used more than once each time. Therefore, each of the N in voice 1 may share the same series of resources.

As in the prior art, the voice 4 produces the required output sample 16 based on the command received at the controller 2 and the interpolation degree of the system.

Often the sound to be produced by a particular voice 4 does not match in frequency with one of the stored sequences of samples. Therefore, the voice 4 must 'shift' the frequency of the stored sequence to make the sound the required frequency.

For example, if the sequence of stored samples represents an intermediate C note on the piano, the sequence can be shifted in frequency to obtain a C # note or a D note.

The frequency of the required sound can be expressed as a multiple of the frequency of the stored sequence. This multiple is represented by the rational number M / L and is known as the phase increment.

Therefore, if the required frequency is twice the frequency of the stored sequence, the phase increase will be equal to two. If the required frequency is half the frequency of the stored sequence, the phase increment will be equal to 1/2. In an example where C # negatives are required, the phase increase will be root 2 (bunch) of 12 minutes which can be accessed by rational numbers.

Often, when the frequency of a sequence of stored samples is shifted, the required samples are not stored in memory. In other words, the required sample falls between two stored samples.

Therefore, the voice 4 retrieves the same number of filter coefficients from the filter table 8 and the number of samples around the required sample from the wave-table memory 6. Each sample retrieved from the wave-table memory 6 is multiplied by the appropriate filter coefficients from the filter table 8, and then the outputs are combined to produce the output 16 of the voice.

If the wave-table memory 6 contains the required samples, the other samples retrieved from the wave-table memory 6 are multiplied by the zero filter coefficients and the coefficients of the filter table 8 are selected so that the stored samples are output. do.

In the preferred embodiment where the filter table 8 includes a value representing the sinc function, the period of the sinc function is twice the sample period.

Each of the voice 4 outputs 16 is transmitted to the mixer 10 where the outputs 16 of all active voices 4 are combined into a combined output 18 and passed to the DAC module 12.

The DAC module 12 includes one or more digital-to-analog converters that convert the combined output 18 of the mixer 10 into an analog signal 20.

2 shows a method performed by the controller 2 of FIG. 1 according to the invention.

In step 101, the controller 2 analyzes the MIDI data stream and determines the number of voices 4 to be active during the next sample period. That is, the controller 2 determines how many voices 4 will supply the output 16 to the mixer 10.

In step 103, the controller 2 determines the number of samples to be used by each voice 4 in calculating the next output 16 (known as the interpolation degree I _D ) and sends it to the voice 4 as appropriate. Instruct.

In step 105, each active voice 4 uses a number of stored samples, such as interpolation degree I _D in the calculation, to calculate the output 16 based on the command received at the controller 2. Each active voice 4 will also use multiple filter coefficients from the filter coefficient table 8, such as interpolation degree I _D.

The sequence repeats for each output cycle, that is, the sequence repeats once per sample period.

In the embodiment of the invention described with reference to FIGS. 3 and 4, the synthesizer has 24 voices 4 and 11 maximum interpolation degrees.

3 is a table illustrating a method of determining an interpolation degree based on the number of active voices according to the present invention. In particular, the table gives the interpolation degree to be used for a given number of active voices.

For example, if the controller 2 determines that only one voice 4 will be active during the next sample period, the controller 2 instructs the voice 4 to use eleven interpolation degrees.

As the number of active voices 4 increases, the interpolation degree used in the calculation of the output 16 decreases linearly.

If all (24) voices 4 of the synthesizer are active, the controller 2 determines that four interpolation degrees will be used.

Alternatively, if the maximum computational complexity is limited for a synthesizer, such as a synthesizer to be used in a portable device, the interpolation degree will be chosen so that the maximum computational complexity is not excessive.

4 is another table illustrating this approach. Again, as the number of active voices 4 increases, the interpolation degree decreases. However, the change is not linear. Instead, the interpolation degree is calculated so that the maximum computational complexity is not excessive.

For example, if a synthesizer has 24 voices, 11 maximum interpolation degrees and consumes 0.5 MIPS / degree / voice (million instruction seconds / degree / voice), a conventional synthesizer would require up to 132 MIPS. This computing power far exceeds that available in common current portable devices such as mobile terminals.

Using the scheme shown in FIG. 4, the computational power will not exceed 50 MIPS. This value is more suitable for portable devices.

The actual manner used will be determined by the computational power available to the synthesizer and the amount of computational power required to implement each interpolation degree.

 5 illustrates the voice of FIG. 1 in more detail. The voice 4 is shown with the controller 2, the wave-table memory 6 and the filter table 8.

Processor 22 receives instructions related to voice 2 from controller 2. The command will include MIDI information related to the voice and an indication related to the interpolation degree to be used in calculating the next output 16.

The controller 2 will indicate to each voice 4 the actual interpolation degree to be used in calculating the next output, or alternatively, the controller 2 will display the number of active voices to each voice 4. The processor 22 will then determine the appropriate interpolation degree.

The processor 22 is connected to a phase increment register 24, a counter 26 and a filter coefficient selector 28,

Filter coefficient selector 28 is connected to filter table 8 to retrieve suitable filter coefficients.

Filter coefficient selector 28 is also connected to counter 26.

In accordance with the present invention, processor 22 informs counter 26 and filter coefficient selector 28 of the interpolation degree to be used to calculate the next output 16.

The processor 22 sets the value of the phase increase register 24 to produce the required output 16. The value of the phase increase amount register 24 is M / L, where L and N are integers and are determined by the processor 22 based on the instructions received from the controller 2.

The phase increase amount value is passed to the adder 30. The adder 30 is connected to a phase register 32 which records the current phase. The output of the adder 30 includes an integer part and a fractional part.

Both the integer part and the fractional part at the output of the phase register are fed back to the adder 30.

The integer portion at the output of the phase register 32 is also passed to a second adder 34 appended to the output of the counter 26. The integer output of the adder 34 is connected to the wave-table memory 6 and determines the sample to be read.

Samples retrieved from the wave-table memory are passed to a multiply-accumulate circuit 36.

In addition to the adder 30, a fractional portion of the output of the phase register 32 is supplied to the filter coefficient selector 28.

The output of the filter coefficient selector 28 is passed to a multiply-accumulate circuit 36, in which the output is combined with the samples retrieved from the wave-table memory 6. .

The operation of the voice 4 is briefly described.

When the input of the phase register 32 is not an integer value, that is, when the fractional part is not zero, the required sample is placed between two tabulated samples. Therefore the required sample must be calculated.

The adder 30 operates once per sample period to add the phase increase amount in the phase increase amount register 24 to the current phase (provided by the phase register 32).

The integer portion of the phase register 32 output represents the wave-table memory address containing the sample already stored before the required sample. In order to calculate the required sample, a number of samples, such as I _D , are read from the wave-table memory 6.

Counter 26 increments by 1 each time an I _D sample is selected around the required sample. Therefore, if I _D is 8, 4 samples before the required sample are read with 4 samples after the required sample. If I _D is 5, three samples before the required sample are read with two samples after the required sample. Alternatively, two samples before the required sample are read and three samples after the required sample are read. These samples are passed to the multiply-accumulate circuit 36.

Note that the counter runs from its initial value to its last value for each sample period.

Filter coefficient selector 28 obtains suitable filter coefficients from filter table 8 according to the interpolation degree with the fractional portion of the phase register output. Filter coefficient selector 28 is controlled by counter 26 to obtain an I _D coefficient from filter table 8.

As soon as filter coefficients 44 are obtained from filter table 8, input received from counter 26 is used to pass filter coefficients to multiply-accumulate circuit 36. Here, the sample obtained from the wave-table memory 6 is multiplied by a suitable filter coefficient 44, and the output is added to obtain an output 16 for the voice.

As the fractional part of the phase register 32 changes, the filter coefficient obtained from the filter table 8 will change.

As the number of active voices 4 changes, the processor will instruct the counter 26 and filter coefficient selector 28 appropriately for the required interpolation degree.

6 illustrates a mobile phone including a music synthesizer in accordance with the present invention. Although the invention is described as being to be incorporated into a mobile phone, the invention is such as a personal digital assistant (PDA), a pager, an electronic organizer or any other desired device that can recreate high quality polyphonic sounds. Any portable device would be considered appropriate.

As conventionally, mobile phone 46 includes an antenna, transceiver circuitry 50, CPU 52, memory 54, and speaker 56.

Mobile phone 46 also includes a MIDI synthesizer 58 in accordance with the present invention. CPU 52 provides MIDI files to MIDI synthesizer 58. The MIDI file will be stored in memory 54 or downloaded from the network via antenna 48 and transceiver circuitry 50.

Thus, a sound synthesizer is described that reduces the computational requirements of the synthesizer, including a high polyphony degree, while ensuring that audible artifacts are minimized.

Claims (16)

In the synthesizer, A memory comprising a plurality of stored samples; Means for calculating an output signal for each of the plurality of active voices, using a plurality of samples selected from the stored samples for each of the plurality of active voices, By means of said calculating means, the number of samples used for each active voice depends on the number of active voices. The method of claim 1, By the means for calculating, the number of samples used for each active voice decreases as the number of active voices increases. The method of claim 2, By the means for calculating, the number of samples used for each active voice is reduced as the number of active voices increases so that the maximum computational complexity is not excessive. The method of claim 1, By means of said calculating means, the number of samples used for each active voice decreases non-linearly as the number of active voices increases. The method according to any one of claims 1 to 4, And said plurality of samples stored in said memory comprises samples of sheet music. The method of claim 5, And said plurality of samples stored in said memory comprises samples of sheet music produced by various electronic musical instruments. The method according to any one of claims 1 to 6, Means for calculating the output signal comprises a filter table. The method of claim 7, wherein The filter table comprises a coefficient of the sinc function. The method according to any one of claims 1 to 8, The synthesizer is a MIDI music synthesizer, characterized in that. In a portable device, 10. A portable device comprising a synthesizer as claimed in claim 1. The method of claim 10, And the portable device is a mobile phone. The method of claim 10, And the portable device is a pager. In a method of operating a synthesizer comprising a plurality of samples stored in a memory, Determining a number of voices to be active in producing a sound; Determining an interpolation degree, based on the number of voices to be activated, limited to the number of samples to be selected from the plurality of samples stored in the memory; And Calculating the output for each active voice using the number of stored samples determined by the interpolation degree. The method of claim 13, And said interpolation degree decreases with increasing number of active voices. The method of claim 13, And said interpolation degree decreases as the number of active voices increases so as not to exceed the maximum computational complexity. The method of claim 13, And said interpolation degree is non-linearly reduced as the number of active voices increases.